Video navigation

Abstract

A system and method for video navigation are disclosed. Motion analysis can be performed upon camera images to determine movement of a vehicle, and consequently present position of the vehicle. Feature points can be identified upon a video image. Movement of the feature points between video frames is indicative of movement of the vehicle. Video navigation can be used, for example, in those instances wherein GPS navigation is unavailable.

I claim: 
     
       1. A method of navigation for an aircraft, the method comprising:
 during a period when a primary GPS navigation system is operating:
 (a) capturing a first image frame and subsequent second image frame of ground features from a video camera mounted on the aircraft, 
 (b) identifying a set of feature points in the first and second image frames by dividing each image frame into grid cells each having a plurality of points, selecting from each grid cell a single point with a largest gradient as the largest gradient point for the corresponding grid cell, and identifying the set of feature points as those largest gradient points that exceed a threshold gradient, 
 (c) determining a change of position of the feature points by how they are mapped from the first image frame to the second image frame including identifying feature points that are moving with respect to a background of the first image frame by determining expected positions of the feature points on the second image frame based upon a position of the feature points in the first image frame and based upon a predetermined motion of the aircraft, wherein moving feature points are removed from the set of identified feature points, 
 (d) within a processor, translating the change of position for the feature points into a video-determined change of position for the aircraft, and 
 estimating bias errors in the video-determined change of position using a GPS-derived change of position to obtain an on-line calibration of the bias errors; and 
 
 during a period of loss of a primary GPS navigation solution:
 repeating acts (a) through (d) to obtain the video-determined change of position, 
 correcting the video-determined change of position according to the bias errors to provide a corrected video-determined change of position, 
 periodically determining position fixes to update the corrected video-determined change of position, and 
 navigating the aircraft by dead reckoning based upon the aircraft's corrected video-determined change of position. 
 
 
     
     
       2. The method of  claim 1 , wherein translating the change of position for the feature points is responsive to an altitude of the aircraft, a camera focal length for the video camera, and a camera pointing angle for the video camera. 
     
     
       3. The method of  claim 2 , wherein the translating the change of position comprises determining a distance from the aircraft to ground positions corresponding to the feature points based on an altitude of the aircraft above ground and the camera point down angle. 
     
     
       4. The method of  claim 3 , further comprising:
 determining terrain heights of the ground positions corresponding to the feature points by looking up a database of terrain elevation; and 
 correcting the distance from the aircraft to the ground positions corresponding to the feature points in response to the terrain heights of the ground positions corresponding to the feature points. 
 
     
     
       5. The method of  claim 4 , wherein the correcting the distance comprises:
 recomputing the ground positions corresponding to the feature points in response to the corrected distance; and 
 repeating the determining of the terrain heights and the correcting of the distance using the recomputed ground positions. 
 
     
     
       6. The method of  claim 3 , wherein the translating the change of position further comprises converting a pixel unit representation of the change of position for the feature points into the video-determined change of position based on the distance from the aircraft to the ground positions corresponding to the feature points. 
     
     
       7. The method as recited in  claim 1 , wherein determining how the feature points are mapped from a first image to a second image comprises using an optical flow algorithm. 
     
     
       8. The method as recited in  claim 1 , wherein determining how the feature points are mapped comprises using an L-K optical flow algorithm with a multi-resolution pyramid. 
     
     
       9. The method of  claim 1 , wherein the position fixes are determined by comparing images of objects on the ground to objects of stored images, wherein positions of the objects of the stored images are known. 
     
     
       10. The method of  claim 1 , wherein the processor is collocated with respect to the video camera. 
     
     
       11. The method of  claim 1 , wherein the processor is located remotely with respect to the video camera. 
     
     
       12. The method of  claim 1 , wherein each image frame is divided into approximately 10 by 15 grid cells.